Weak homotopy equivalences #
A continuous map f : X → Y is a weak homotopy equivalence if it induces a bijection on path
components and a bijection π_n(X, x) → π_n(Y, f x) for every n ≥ 1 and every point x : X
(Hatcher, Algebraic Topology, Section 4.1). This file defines the predicate
ContinuousMap.IsWeakHomotopyEquivalence on Mathlib's cubical homotopy groups and proves its
closure properties: homotopy equivalences are weak homotopy equivalences, the predicate depends
only on the homotopy class of the map, and it is closed under composition, under left and right
cancellation, and under retracts. In a commuting square whose two parallel maps are weak homotopy
equivalences, the other two maps have the same connectivity.
Whitehead's theorem, that a weak homotopy equivalence between spaces of CW type is a homotopy
equivalence, is ContinuousMap.IsWeakHomotopyEquivalence.homotopyEquiv in
TauCeti.AlgebraicTopology.Homotopy.Whitehead.
Main declarations #
ContinuousMap.IsWeakHomotopyEquivalence: the predicate.ContinuousMap.IsWeakHomotopyEquivalence.map_bijective: the induced map on the homotopy group indexed by any finite type of cube directions is bijective.ContinuousMap.IsWeakHomotopyEquivalence.joined_iff: a weak homotopy equivalence is injective on path components.ContinuousMap.IsWeakHomotopyEquivalence.mapOfEq_bijective: a weak homotopy equivalence induces bijections of fundamental groups.ContinuousMap.isWeakHomotopyEquivalence_iff_forall_bijective: on a path-connected space it suffices to check bijectivity on homotopy groups at a single base point.HomotopyGroup.map_injective_of_isEmpty_of_joined,HomotopyGroup.map_surjective_of_isEmpty_of_exists_joined: in degree zero, the induced map is injective or surjective as soon as the map is on path components.ContinuousMap.isWeakHomotopyEquivalence_iff_joined: a weak homotopy equivalence is a bijection on path components inducing bijections on the positive-dimensional homotopy groups.ContinuousMap.HomotopyEquiv.isWeakHomotopyEquivalence: a homotopy equivalence is a weak homotopy equivalence.ContinuousMap.IsWeakHomotopyEquivalence.of_homotopic,ContinuousMap.IsWeakHomotopyEquivalence.comp,ContinuousMap.IsWeakHomotopyEquivalence.of_postcomp,ContinuousMap.IsWeakHomotopyEquivalence.of_precomp: invariance under homotopy, composition, and left and right cancellation (two out of three).ContinuousMap.IsWeakHomotopyEquivalence.of_retract: a retract of a weak homotopy equivalence is a weak homotopy equivalence.ContinuousMap.isWeakHomotopyEquivalence_of_contractibleSpace: every map between contractible spaces is a weak homotopy equivalence.ContinuousMap.IsWeakHomotopyEquivalence.forall_surjective_map_iff_of_comp_eq,ContinuousMap.IsWeakHomotopyEquivalence.forall_injective_map_iff_of_comp_eq: in a squareg ∘ i = j ∘ fwithfandgweak homotopy equivalences,jinduces surjections (injections) on a homotopy group at every base point exactly whenidoes.
References #
- A. Hatcher, Algebraic Topology, Section 4.1, the definition of weak homotopy equivalence preceding Theorem 4.5.
Injectivity in degree zero through path components. For an empty type N of cube
directions, π_N(X, x) is the set of path components of X, so the map induced by f is
injective when two points of X are joined whenever their images are.
Surjectivity in degree zero through path components. For an empty type N of cube
directions, π_N(Y, y) is the set of path components of Y, so the map induced by f is
surjective when every point of Y is joined to a point of the image of f.
A continuous map f : X → Y is a weak homotopy equivalence if every point of Y is
joined by a path to a point of the image of f, and f induces a bijection
π_n(X, x) → π_n(Y, f x) for every n and every x : X.
Since π_0(X, x) is the set of path components of X, the condition in degree 0 says that f
is injective on path components, and also surjective on them as soon as X is nonempty. The
field exists_joined, surjectivity on path components, adds to this only when X is empty: it
then forces Y to be empty as well.
- exists_joined (y : Y) : ∃ (x : X), Joined (f x) y
Every point of the target is joined by a path to a point of the image.
The induced map
π_n(X, x) → π_n(Y, f x)is bijective for everynand everyx.
Instances For
A weak homotopy equivalence induces a bijection on the homotopy groups indexed by any finite
type of cube directions, not only on the groups π_n indexed by Fin n.
A weak homotopy equivalence is injective on path components: two points of X are joined by
a path exactly when their images are.
A map homotopic to a weak homotopy equivalence is a weak homotopy equivalence.
The identity map is a weak homotopy equivalence.
A composite of weak homotopy equivalences is a weak homotopy equivalence.
Two out of three, left cancellation. If g and g ∘ f are weak homotopy equivalences,
then so is f.
Two out of three, right cancellation. If f and g ∘ f are weak homotopy equivalences,
then so is g. At a point y of Y, the induced map on homotopy groups is compared, by
base-point change along a path from a point f x to y and along its image under g, with the
induced map at f x, which is bijective by cancellation on the right.
Retracts. A retract of a weak homotopy equivalence is a weak homotopy equivalence: if
f : X → Y is a retract of g : X' → Y', through i : X → X' and j : Y → Y' with retractions
r and s such that g ∘ i = j ∘ f and f ∘ r = s ∘ g, and g is a weak homotopy equivalence,
then so is f. On every homotopy group the map induced by f is a retract of the map induced by
g, and a retract of a bijection is a bijection.
A weak homotopy equivalence induces a bijection of fundamental groups at every base point.
Squares of weak homotopy equivalences #
Let g ∘ i = j ∘ f for maps i : X → X' and j : Y → Y', where f : X → Y and g : X' → Y'
are weak homotopy equivalences. Then i and j have the same connectivity: each of the
following properties holds for i exactly when it holds for j. Every point of the target is
joined by a path to a point of the image, and the induced map on a given homotopy group is
surjective, or injective, at every base point.
In a square g ∘ i = j ∘ f whose maps f and g are weak homotopy equivalences, every point
of Y' is joined to a point of the image of j exactly when every point of X' is joined to a
point of the image of i.
In a square g ∘ i = j ∘ f whose maps f and g are weak homotopy equivalences, the map
π_N(Y, f x) → π_N(Y', j (f x)) induced by j is surjective exactly when the map
π_N(X, x) → π_N(X', i x) induced by i is.
In a square g ∘ i = j ∘ f whose maps f and g are weak homotopy equivalences, the map
π_N(Y, f x) → π_N(Y', j (f x)) induced by j is injective exactly when the map
π_N(X, x) → π_N(X', i x) induced by i is.
In a square g ∘ i = j ∘ f whose maps f and g are weak homotopy equivalences, j induces
surjections π_N(Y, y) → π_N(Y', j y) at every point y exactly when i induces surjections
π_N(X, x) → π_N(X', i x) at every point x.
In a square g ∘ i = j ∘ f whose maps f and g are weak homotopy equivalences, j induces
injections π_N(Y, y) → π_N(Y', j y) at every point y exactly when i induces injections
π_N(X, x) → π_N(X', i x) at every point x.
On a path-connected space, weak homotopy equivalences can be detected at one point. A
map out of a path-connected space is a weak homotopy equivalence exactly when it induces a
bijection π_n(X, x) → π_n(Y, f x) for every n at a single base point x. In degree 0 this
says that Y is path-connected as well.
Weak homotopy equivalences through path components. A map is a weak homotopy equivalence
exactly when it induces a bijection on path components, that is, every point of Y is joined to a
point of the image and two points of X are joined when their images are, and a bijection
π_(n+1)(X, x) → π_(n+1)(Y, f x) for every n and every point x.
A homotopy equivalence is a weak homotopy equivalence.
A map between contractible spaces is a weak homotopy equivalence. It is homotopic to a homotopy equivalence, since any two maps into a contractible space are homotopic.